I am surprised at how tracking the results of ordinary traders can achieve profits, as a trader's behavior indicates loss on average, otherwise many would not have made a fortune, so all trading groups focus on copying the trading of experienced users, or you would prefer the analysis of an expert in stocks to that of your friend.
So this isn't really geared towards specific traders' thoughts, as in you would see one or two people's chats and make a decision on that. It's more or less to show an aggregate of many different sentiments all at once. Let's say you saw a lot of trades of a particular token happening in the chat or a lot of people having genuine conversation about a token, you maybe could use that information to make a buy or sell decision. It's not to replace expert analysis by any means but since the market is dictated by the collection of individual users, knowing the "average" of what they think as a whole could be useful.
How will you ensure that abuse is prevented, since the pumping team can register hundreds of accounts and fool the artificial intelligence that there is activity for traders to buy a specific pair?
I have already implemented methods to stop spamming of my API and to the chat. If you manually try to send more than a few chats per second right now the server will stop you. This alone will not stop pumping but there are a few methods I am working towards right now such as chat voting. Chat voting would basically mean that users can upvote or downvote chats, which will greatly affect their gravity on the AI. This can also be based on the user where if a user gets a certain amount of downvotes in a certain amount of time, then their chats' effect on the AI can be nullified. Methods like these plus IP tracking and a reputation system based on IP/user can be implemented as well.
Will you be able to display the trading results on the platform and the real-time prices so we can compare them?
Absolutely, real-time pricing exists right now if you select a pair and type in a value to trade. However in the future I plan on recording trade data and showing those statistics. Because the exchange uses a liquidity aggregator I will probably use an API like bitquery to query the pricing data of the entire aggregator to show this data in the nearer future until the volume of Chatswap is large enough to show enough data.
Isn't that a bad idea though? As far as I can see, people failed to make profits due to a failure to keep their emotions in check. Whether they cut losses too early or take profit too early even though technical analysis shows otherwise. They also failed to stick to their target and didn't manage their risk properly based on their risk profile.
You are definitely correct, people make bad decisions based on their emotions. Despite this they still make decisions that affect the market. This data that would be collected wouldn't necessarily be used to follow what others are doing, if you see a large irrational sentiment and trading decisions towards a token you can choose to make the opposite decision. The goal of this exchange would be to see why people are making the trades they are in real-time, but there's no set use for this it's simply showing data.
Not to mention your current alpha is just a chatbox with a trading window, like what other exchanges already have today.
Definitely a very early stage prohect but one difference I've already implemented is seeing real-time trades in chat. If you make a trade you will be able to see it in chat.